import os import re import sys import json from collections import Counter # Ensure UTF-8 output on Windows sys.stdout.reconfigure(encoding='utf-8') tap01_path = r"c:\Projects\KieuStory\FilmMaker\TAP_01_XUAN_SAC_THE_NGUYEN_VA_GIONG_BAO_DOAN_TRUONG.md" idx_path = r"c:\Projects\KieuStory\FilmMaker\INDEX_VA_DANH_MUC_CANH_QUAY.md" sb_path = r"c:\Projects\KieuStory\03_Storyboards\Ep01_Gia_Bien_Storyboard.md" epguide_path = r"c:\Projects\KieuStory\01_Scripts_And_Episodes\EPISODE_GUIDE_30MIN.md" with open(tap01_path, 'r', encoding='utf-8') as f: tap01 = f.read() with open(idx_path, 'r', encoding='utf-8') as f: idx = f.read() with open(sb_path, 'r', encoding='utf-8') as f: sb = f.read() with open(epguide_path, 'r', encoding='utf-8') as f: epguide = f.read() # Pattern matching shot headings (5 or 6 hashes) shot_pattern = re.compile(r'#+\s+Shot\s+(\d+):\s+`?(ep01_scene\d+_shot\d+)`?\s+\(10s\)') matches = list(shot_pattern.finditer(tap01)) shots = [] for i, m in enumerate(matches): start = m.start() end = matches[i+1].start() if i+1 < len(matches) else len(tap01) text = tap01[start:end] shot_num = int(m.group(1)) shot_id = m.group(2) # Check sections has_cam = ('- **Cỡ cảnh & Máy quay**:' in text) or ('**Cỡ cảnh' in text) has_vis = ('- **Thị giác**:' in text) or ('**Thị giác' in text) has_aud = ('- **Âm thanh 4 Stems**:' in text) or ('**Âm thanh' in text) has_stem1 = '*Stem 1' in text has_stem2 = '*Stem 2' in text has_stem3 = '*Stem 3' in text has_stem4 = '*Stem 4' in text has_thoai = ('- **Thoại & Khẩu hình**:' in text) shots.append({ 'shot_id': shot_id, 'shot_num': shot_num, 'char_len': len(text.strip()), 'has_cam': has_cam, 'has_vis': has_vis, 'has_aud': has_aud, 'has_all_4_stems': has_stem1 and has_stem2 and has_stem3 and has_stem4, 'has_thoai': has_thoai, 'text': text }) # Compute statistics char_lengths = [s['char_len'] for s in shots] min_len = min(char_lengths) max_len = max(char_lengths) avg_len = sum(char_lengths) / len(char_lengths) # Missing components missing_cam = [s['shot_id'] for s in shots if not s['has_cam']] missing_vis = [s['shot_id'] for s in shots if not s['has_vis']] missing_aud = [s['shot_id'] for s in shots if not s['has_aud']] missing_stems = [s['shot_id'] for s in shots if not s['has_all_4_stems']] # Duplicate shot texts check (near duplicates) visual_texts = [] for s in shots: # extract visual text vis_match = re.search(r'- \*\*Thị giác\*\*:(.*?)(?=- \*\*|\n\n|#####|$)', s['text'], re.DOTALL) if vis_match: visual_texts.append((s['shot_id'], vis_match.group(1).strip())) else: visual_texts.append((s['shot_id'], '')) vis_text_counts = Counter(t[1] for t in visual_texts) duplicated_visuals = [t for t, count in vis_text_counts.items() if count > 1 and len(t) > 20] # Scene Breakdown scene_counts = {} for s in shots: sc = int(re.search(r'ep01_scene(\d+)_', s['shot_id']).group(1)) scene_counts[sc] = scene_counts.get(sc, 0) + 1 act1 = sum(scene_counts[i] for i in range(1, 5)) act2 = sum(scene_counts[i] for i in range(5, 10)) act3 = sum(scene_counts[i] for i in range(10, 16)) # Check placeholders placeholders = re.findall(r'\b(TODO|TBD|FIXME|PLACEHOLDER|XXX|đang cập nhật|tự code|như trên)\b', tap01, re.I) # Check male crying # Check occurrences of crying words associated with male characters male_names = ["Kim Trọng", "Vương Quan", "Vương Ông", "Từ Hải"] crying_words = ["khóc", "rơi lệ", "nước mắt", "rưng rưng", "giọt lệ", "nghẹn ngào rơi", "khóc than"] male_violations = [] for line_no, line in enumerate(tap01.splitlines(), 1): for male in male_names: if male in line: for cw in crying_words: if cw in line: line_lower = line.lower() # if negated or explicitly about female / object if any(neg in line_lower for neg in ["không", "triệt tiêu", "0%", "nén", "chối từ", "bảo vệ"]): continue if "của kiều" in line_lower or "thúy kiều" in line_lower or "vương bà" in line_lower: # check if male is just mentioned e.g. "trao cho Kim Trọng chiếc quạt... nước mắt Kiều" if re.search(r'(kiều|vương bà)[^.\n]*?(nước mắt|rơi lệ|giọt lệ|khóc)', line_lower) or \ re.search(r'(nước mắt|giọt lệ)[^.\n]*?(kiều|của kiều)', line_lower): continue male_violations.append((line_no, male, cw, line.strip())) # Output Report report = { "total_shots": len(shots), "unique_shot_ids": len(set(s['shot_id'] for s in shots)), "char_length_stats": { "min": min_len, "max": max_len, "avg": round(avg_len, 1) }, "completeness": { "missing_camera": missing_cam, "missing_visuals": missing_vis, "missing_audio": missing_aud, "missing_4_stems": missing_stems }, "duplicated_visual_descriptions": len(duplicated_visuals), "placeholders_found": placeholders, "scene_shot_counts": scene_counts, "act_breakdown": { "Act 1 (Scenes 01-04)": act1, "Act 2 (Scenes 05-09)": act2, "Act 3 (Scenes 10-15)": act3, "Total": act1 + act2 + act3 }, "male_tears_violations": male_violations, "dam_tien_sequence": { "scene05_shots": scene_counts.get(5, 0), "C05-A_shots": len([s for s in shots if 'scene05_' in s['shot_id'] and s['shot_num'] <= 8]), "C05-B_shots": len([s for s in shots if 'scene05_' in s['shot_id'] and 9 <= s['shot_num'] <= 15]), "C05-C_shots": len([s for s in shots if 'scene05_' in s['shot_id'] and 16 <= s['shot_num'] <= 27]), "has_dau_hai_in_reu": "dấu hài" in tap01.lower() and "in rêu" in tap01.lower(), "has_tram_bac_khac_tho": "trâm bạc" in tap01.lower() and "vạch da cây" in tap01.lower() }, "lip_sync_guard_shot08": "[KHÓA KHẨU HÌNH BẮT BUỘC R5]" in [s['text'] for s in shots if s['shot_id'] == 'ep01_scene08_shot08'][0], "storyboard_check": "tiếng Vương Ông kêu nghẹn ngào" not in sb, "epguide_check": "Tiếng khóc than xé lòng của gia đình Vương viên ngoại" not in epguide, "index_summary_verified": ("15 Phân cảnh | 188 Shots | Thời lượng: 29 phút 40 giây" in idx) and ("56 Phân cảnh điện ảnh lớn" in idx) } print(json.dumps(report, indent=2, ensure_ascii=False))